PHP Zend_Search_Lucene_Field::Keyword方法代码示例

 function index()
 {
     $this->load->library('zend', 'Zend/Feed');
     $this->load->library('zend', 'Zend/Search/Lucene');
     $this->load->library('zend');
     $this->zend->load('Zend/Feed');
     $this->zend->load('Zend/Search/Lucene');
     //Create index.
     $index = new Zend_Search_Lucene('C:\\xampp\\xampp\\htdocs\\controle_frota\\lucene\\feeds_index', true);
     $feeds = array('http://oglobo.globo.com/rss.xml?limite=50');
     //grab each feed.
     foreach ($feeds as $feed) {
         $channel = Zend_Feed::import($feed);
         echo $channel->title() . '<br />';
         //index each item.
         foreach ($channel->items as $item) {
             if ($item->link() && $item->title() && $item->description()) {
                 //create an index doc.
                 $doc = new Zend_Search_Lucene_Document();
                 $doc->addField(Zend_Search_Lucene_Field::Keyword('link', $this->sanitize($item->link())));
                 $doc->addField(Zend_Search_Lucene_Field::Text('title', $this->sanitize($item->title())));
                 $doc->addField(Zend_Search_Lucene_Field::Unstored('contents', $this->sanitize($item->description())));
                 echo "\tAdding: " . $item->title() . '<br />';
                 $index->addDocument($doc);
             }
         }
     }
     $index->commit();
     echo $index->count() . ' Documents indexed.<br />';
 }

 private function __construct($rule, $additional, $storeContent)
 {
     /*$dom = new DOMDocument ();
     		$dom->preserveWhiteSpace = false;
     		$dom->loadXML ( $rule );*/
     //walk through the association rule and index it
     //$this->_parse ( $rule, 1 );
     $quantifiers = $rule->childNodes;
     foreach ($quantifiers as $quantifier) {
         if ($quantifier->nodeName == '#text' || $quantifier->nodeValue == '') {
             continue;
         }
         //print $quantifier->nodeName.' : '. trim($quantifier->nodeValue).'<br/>';
         $val = trim($quantifier->nodeValue);
         if (is_numeric($val)) {
             $val = JuceneHelper::prepareNumber($val);
         } else {
             $val = (string) $val;
             $val = str_replace("-", "", $val);
         }
         if ($quantifier->nodeName == 'Text') {
             $type = 'Unindexed';
         } else {
             $type = 'Text';
         }
         $this->addField(Zend_Search_Lucene_Field::$type($quantifier->nodeName, $val, JUCENE_ENCODING));
     }
     foreach ($additional as $field => $value) {
         if (is_numeric($value)) {
             $val = JuceneHelper::prepareNumber($value);
         }
         $this->addField(Zend_Search_Lucene_Field::Keyword('service_' . $field, $value, JUCENE_ENCODING));
     }
 }

 public function IndexBug($bug)
 {
     $this->RemoveBug($bug->bug_id);
     $doc = new Zend_Search_Lucene_Document();
     $doc->AddField(Zend_Search_Lucene_Field::Keyword('bug_id', $bug->bug_id));
     $doc->AddField(Zend_Search_Lucene_Field::Text('title', $bug->title));
     $doc->AddField(Zend_Search_Lucene_Field::Keyword('reporting_user_id', $bug->reporting_user_id));
     $doc->AddField(Zend_Search_Lucene_Field::Keyword('reporting_date', $bug->reporting_date));
     // We concatenate all comments into a single text blob. We only show
     // hits as bugs, but we want comment content to matter.
     $comment_blob = '';
     $stmt = Bugdar::$db->Prepare("SELECT body FROM " . TABLE_PREFIX . "comments WHERE bug_id = ? ORDER BY comment_id");
     $stmt->Execute(array($bug->bug_id));
     while ($comment = $stmt->FetchObject()) {
         $comment_blob .= $comment->body . "\n\n";
     }
     $doc->AddField(Zend_Search_Lucene_Field::UnStored('comments', $comment_blob));
     // Add all attributes.
     $stmt = Bugdar::$db->Prepare("SELECT * FROM " . TABLE_PREFIX . "bug_attributes WHERE bug_id = ?");
     $stmt->Execute(array($bug->bug_id));
     $tags = array();
     while ($attr = $stmt->FetchObject()) {
         if ($attr->attribute_title) {
             $doc->AddField(Zend_Search_Lucene_Field::Keyword($attr->attribute_title, $attr->value));
         } else {
             $tags[] = $attr->value;
         }
     }
     $doc->AddField(Zend_Search_Lucene_Field::Text('tag', implode(' ', $tags)));
     $this->lucene->AddDocument($doc);
 }

 public function updateLuceneIndex()
 {
     //delete existing entries
     $index = $this->getTable()->getLuceneIndex();
     // remove existing entries
     foreach ($index->find('pk:' . $this->getId()) as $hit) {
         $index->delete($hit->id);
     }
     // create new Lucene document
     $doc = new Zend_Search_Lucene_Document();
     // store product primary key to identify it in the search results
     $doc->addField(Zend_Search_Lucene_Field::Keyword('pk', $this->getId()));
     $tr = Doctrine::getTable('ProductTranslation')->createQuery()->from('ProductTranslation pt')->where('pt.id = ?', $this->getId())->execute();
     $doc->addField(Zend_Search_Lucene_Field::UnStored('original_title', $this->getOriginalTitle(), 'utf-8'));
     // add fields to index depending on existing Translations
     foreach ($tr->toArray() as $transArr) {
         $lang = $transArr['lang'];
         unset($transArr['lang'], $transArr['id'], $transArr['volume'], $transArr['slug']);
         foreach ($transArr as $field => $value) {
             $fieldName = $field . '_' . $lang;
             // (name_en, name_fi),  (description_en, description_fi)
             $doc->addField(Zend_Search_Lucene_Field::UnStored($fieldName, strip_tags($value), 'utf-8'));
         }
     }
     // add product to the index
     $index->addDocument($doc);
     $index->commit();
 }

 public function edit($needFields = array(), $data = array(), $charset = 'UTF-8')
 {
     $index = new Zend_Search_Lucene(ZY_ROOT . '/index');
     $doc = new Zend_Search_Lucene_Document();
     foreach ($needFields as $key => $field) {
         switch ($field) {
             case 'keywords':
                 $doc->addField(Zend_Search_Lucene_Field::Keyword($key, $data[$key], $charset));
                 break;
             case 'text':
                 $doc->addField(Zend_Search_Lucene_Field::Text($key, $data[$key], $charset));
                 break;
             case 'unindexed':
                 $doc->addField(Zend_Search_Lucene_Field::unindexed($key, $data[$key], $charset));
                 break;
             default:
                 $doc->addField(Zend_Search_Lucene_Field::$field($key, $data[$key], $charset));
                 break;
         }
     }
     $index->addDocument($doc);
     $index->commit();
     $index->optimize();
     return TRUE;
 }

function index_lucene($article, $optimise)
{
    $index = getIndex_lucene();
    $term = new Zend_Search_Lucene_Index_Term($article["PMID"], 'PMID');
    // a pre-existing page cannot be updated, it has to be
    // deleted, and indexed again:
    $exactSearchQuery = new Zend_Search_Lucene_Search_Query_Term($term);
    $hits = $index->find($exactSearchQuery);
    if (count($hits) > 0) {
        echo "[deleting previous version]\n";
        foreach ($hits as $hit) {
            $index->delete($hit->id);
        }
    }
    $doc = new Zend_Search_Lucene_Document();
    $doc->addField(Zend_Search_Lucene_Field::Keyword('PMID', $article["PMID"]));
    $doc->addField(Zend_Search_Lucene_Field::Keyword('Year', $article["Year"]));
    $doc->addField(Zend_Search_Lucene_Field::Keyword('Journal', $article["Journal"]));
    $doc->addField(Zend_Search_Lucene_Field::Text('Title', $article["Title"], 'utf-8'));
    $doc->addField(Zend_Search_Lucene_Field::Text('Authors', $article["Authors"], 'utf-8'));
    $doc->addField(Zend_Search_Lucene_Field::Text('Reference', $article["Reference"], 'utf-8'));
    $doc->addField(Zend_Search_Lucene_Field::UnStored('Abstract', $article["Abstract"], 'utf-8'));
    $doc->addField(Zend_Search_Lucene_Field::Text('MeshHeadings', $article["MeshHeadings"], 'utf-8'));
    $index->addDocument($doc);
    if ($optimise) {
        echo "Optimising index\n";
        $index->optimize();
    }
    $index->commit();
    echo "The index contains " . $index->numDocs() . " documents\n";
}

 public static function update($data)
 {
     try {
         //Update an index.
         $index = Zend_Search_Lucene::open('../application/searchindex');
         Zend_Search_Lucene_Analysis_Analyzer::setDefault(new Zend_Search_Lucene_Analysis_Analyzer_Common_Utf8_CaseInsensitive());
     } catch (Zend_Search_Exception $e) {
         throw $e;
     }
     // remove an existing entry
     $hits = $index->find('pk:' . $data['pk']);
     foreach ($hits as $hit) {
         $index->delete($hit->id);
     }
     $doc = new Zend_Search_Lucene_Document();
     $doc->addField(Zend_Search_Lucene_Field::Keyword('pk', $data['pk']));
     $doc->addField(Zend_Search_Lucene_Field::Keyword('code', $data['code'], 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::Keyword('u_code', strtolower($data['code']), 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::unIndexed('type', $data['type'], 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::unIndexed('id', $data['id'], 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::Text('title', $data['title'], 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::Text('en_title', Default_Model_Functions::convert_vi_to_en($data['title']), 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::Text('description', $data['description'], 'UTF-8'));
     $doc->addField(Zend_Search_Lucene_Field::Text('en_description', Default_Model_Functions::convert_vi_to_en($data['description']), 'UTF-8'));
     $index->addDocument($doc);
     $index->commit();
 }

    /**
     * Construct a Zend_Search_Lucene_Document object out of a document db row.
     * 
     * @global string $urlServer
     * @param  object  $docu
     * @return Zend_Search_Lucene_Document
     */
    protected function makeDoc($docu) {
        global $urlServer;
        $encoding = 'utf-8';

        $doc = new Zend_Search_Lucene_Document();
        $doc->addField(Zend_Search_Lucene_Field::Keyword('pk', 'doc_' . $docu->id, $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Keyword('pkid', $docu->id, $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Keyword('doctype', 'doc', $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Keyword('courseid', $docu->course_id, $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('title', Indexer::phonetics($docu->title), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('content', Indexer::phonetics($docu->description), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('filename', Indexer::phonetics($docu->filename), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('comment', Indexer::phonetics($docu->comment), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('creator', Indexer::phonetics($docu->creator), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('subject', Indexer::phonetics($docu->subject), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('author', Indexer::phonetics($docu->author), $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('visible', $docu->visible, $encoding));
        $doc->addField(Zend_Search_Lucene_Field::Text('public', $docu->public, $encoding));

        $urlAction = ($docu->format == '.dir') ? 'openDir' : 'download';
        $doc->addField(Zend_Search_Lucene_Field::UnIndexed('url', $urlServer
                        . 'modules/document/index.php?course=' . course_id_to_code($docu->course_id)
                        . '&' . $urlAction . '=' . $docu->path, $encoding));

        return $doc;
    }

 /**
  * Constructor. Creates our indexable document and adds all
  * necessary fields to it using the passed in document
  */
 public function __construct($document)
 {
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('id_entry', $document->getId()));
     $this->addField(Zend_Search_Lucene_Field::Keyword('url', $document->getUrl()));
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('creation_date', $document->getCreationDate()));
     $this->addField(Zend_Search_Lucene_Field::Text('name', $document->getName()), 'utf-8');
     $this->addField(Zend_Search_Lucene_Field::Text('content', $document->getDetails()));
     $this->addField(Zend_Search_Lucene_Field::Text('tag', $document->getImplodedTags()));
 }

 /**
  * Updates the index for an object
  *
  * @param Doctrine_Record $object
  */
 public function updateIndex(Doctrine_Record $object, $delete = false)
 {
     /* error checking */
     if (!array_key_exists('models', $this->config) || empty($this->config['models'])) {
         throw new Exception(sprintf('No models set in search.yml', $name));
     }
     if (!array_key_exists($model = get_class($object), $this->config['models'])) {
         throw new Exception(sprintf('Model "%s" not defined in "%s" index in your search.yml', $model, $this->name));
     }
     $id = $this->generateId($object->getId(), $model);
     $config = $this->config['models'][$model];
     //delete existing entries
     foreach ($this->search('_id:"' . $id . '"') as $hit) {
         $this->getIndex()->delete($hit->id);
     }
     if ($delete) {
         return;
     }
     //only add to search if canSearch method on model returns true (search if no method exists)
     if (method_exists($object, 'canSearch')) {
         if (!call_user_func(array($object, 'canSearch'))) {
             return;
         }
     }
     $doc = new Zend_Search_Lucene_Document();
     // store a key for deleting in future
     $doc->addField(Zend_Search_Lucene_Field::Keyword('_id', $id));
     // store job primary key and model name to identify it in the search results
     $doc->addField(Zend_Search_Lucene_Field::Keyword('_pk', $object->getId()));
     $doc->addField(Zend_Search_Lucene_Field::Keyword('_model', $model));
     // store title - used for search result title
     if (!array_key_exists('title', $config)) {
         throw new Exception(sprintf('A title must be set for model "%s" in search.yml', $model));
     }
     $doc->addField(Zend_Search_Lucene_Field::unIndexed('_title', call_user_func(array($object, 'get' . sfInflector::camelize($config['title'])))));
     // store description - used for search result description
     if (!array_key_exists('description', $config)) {
         throw new Exception(sprintf('A description must be set for model "%s" in search.yml', $model));
     }
     $doc->addField(Zend_Search_Lucene_Field::unIndexed('_description', call_user_func(array($object, 'get' . sfInflector::camelize($config['description'])))));
     // store url - @todo add more routing options
     if (!array_key_exists('route', $config)) {
         throw new Exception(sprintf('A route must be set for model "%s" in search.yml', $model));
     }
     sfContext::getInstance()->getConfiguration()->loadHelpers('Url');
     $url = url_for($config['route'], $object);
     $doc->addField(Zend_Search_Lucene_Field::unIndexed('_url', $url));
     //store fields
     if (array_key_exists('fields', $config)) {
         foreach ($config['fields'] as $field => $config) {
             $doc->addField(Zend_Search_Lucene_Field::UnStored($field, call_user_func(array($object, 'get' . sfInflector::camelize($field))), 'utf-8'));
         }
     }
     //save index
     $this->getIndex()->addDocument($doc);
     $this->getIndex()->commit();
 }

 public function insertFromDB($since)
 {
     $doc = new Zend_Search_Lucene_Document();
     $doc->addField(Zend_Search_Lucene_Field::Keyword('id', $row['id']));
     $doc->addField(Zend_Search_Lucene_Field::Keyword('person', $row['person']));
     $doc->addField(Zend_Search_Lucene_Field::Text('celeb_type', $row['celeb_type']));
     $doc->addField(Zend_Search_Lucene_Field::Text('wikikeyword', $row['wikikeyword']));
     $doc->addField(Zend_Search_Lucene_Field::Text('blurb', $row['blurb']));
     $this->search->addDocument($doc);
 }

 /**
  * Index entity of document with given id.
  * 
  * @param Mage_Core_Model_Abstract
  *
  * @return Mage_Lucene_Model_Index_Document_Abstract
  **/
 public function index($sourceModel)
 {
     $this->_sourceModel = $sourceModel;
     $this->_id = $sourceModel->getId();
     $this->delete();
     $this->addField(Zend_Search_Lucene_Field::Keyword('doctype', $this->getDoctype()));
     $this->addField(Zend_Search_Lucene_Field::Keyword('entity_id', $this->_id));
     $this->addField(Zend_Search_Lucene_Field::Keyword(self::STORE_ATTRIBUTE_CODE, $this->getStore()->getId()));
     $this->addAttributes();
     $this->addDocument();
     return $this;
 }

 public function testKeyword()
 {
     $field = Zend_Search_Lucene_Field::Keyword('field', 'value');
     $this->assertEquals($field->boost, 1);
     $this->assertEquals($field->encoding, '');
     $this->assertEquals($field->isBinary, false);
     $this->assertEquals($field->isIndexed, true);
     $this->assertEquals($field->isStored, true);
     $this->assertEquals($field->isTokenized, false);
     $this->assertEquals($field->name, 'field');
     $this->assertEquals($field->value, 'value');
 }

 /**
  * rebuild the index
  *
  * @access public
  * @return void
  */
 function build_index()
 {
     $index = $this->__open(true);
     $index->setMergeFactor(2000);
     $index->setMaxBufferedDocs(500);
     $start = time();
     foreach ($this->settings as $model => $model_options) {
         App::import('Model', $model);
         $model = new $model();
         if (empty($model_options['find_options'])) {
             $model_options['find_options'] = array();
         }
         if (method_exists($model, 'find_index')) {
             $results = $model->find_index('all', $model_options['find_options']);
         } else {
             $results = $model->find('all', $model_options['find_options']);
         }
         if (Configure::read()) {
             $this->log($model->name . ' find time: ' . (time() - $start), 'searchable');
             $start = time();
         }
         $count = count($results);
         $i = 1;
         foreach ($results as $result) {
             printf("%.1f", $i / $count * 100);
             $this->out("");
             $i++;
             $this->out('Processing ' . $model->name . ' #' . $result[$model->name]['id']);
             $doc = new Zend_Search_Lucene_Document();
             // add the model field
             $doc->addField(Zend_Search_Lucene_Field::Keyword('cake_model', $model->name, 'utf-8'));
             foreach ($model_options['fields'] as $field_name => $options) {
                 if (!empty($options['prepare']) && function_exists($options['prepare'])) {
                     $result[$model->name][$field_name] = call_user_func($options['prepare'], $result[$model->name][$field_name]);
                 }
                 $alias = !empty($options['alias']) ? $options['alias'] : $field_name;
                 $doc->addField(Zend_Search_Lucene_Field::$options['type']($alias, $result[$model->name][$field_name], 'utf-8'));
             }
             $index->addDocument($doc);
             $this->out('Processed ' . $model->name . ' #' . $result[$model->name]['id']);
         }
         if (Configure::read()) {
             $this->log($model->name . ' adding time: ' . (time() - $start), 'searchable');
             $start = time();
         }
     }
     $this->optimize($index);
     $index->commit();
     if (Configure::read()) {
         $this->log('Optimize+commit time: ' . (time() - $start));
     }
 }

 public function __construct(&$doc, &$data, $course_id, $group_id, $user_id, $path, $additional_keyset = null)
 {
     $encoding = 'UTF-8';
     //document identification and indexing
     $this->addField(Zend_Search_Lucene_Field::Keyword('docid', $doc->docid, $encoding));
     //document type : the name of the Moodle element that manages it
     $this->addField(Zend_Search_Lucene_Field::Keyword('doctype', $doc->documenttype, $encoding));
     //allows subclassing information from complex modules.
     $this->addField(Zend_Search_Lucene_Field::Keyword('itemtype', $doc->itemtype, $encoding));
     //caches the course context.
     $this->addField(Zend_Search_Lucene_Field::Keyword('course_id', $course_id, $encoding));
     //caches the originator's group.
     $this->addField(Zend_Search_Lucene_Field::Keyword('group_id', $group_id, $encoding));
     //caches the originator if any
     $this->addField(Zend_Search_Lucene_Field::Keyword('user_id', $user_id, $encoding));
     // caches the context of this information. i-e, the context in which this information
     // is being produced/attached. Speeds up the "check for access" process as context in
     // which the information resides (a course, a module, a block, the site) is stable.
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('context_id', $doc->contextid, $encoding));
     //data for document
     $this->addField(Zend_Search_Lucene_Field::Text('title', $doc->title, $encoding));
     $this->addField(Zend_Search_Lucene_Field::Text('author', $doc->author, $encoding));
     $this->addField(Zend_Search_Lucene_Field::UnStored('contents', $doc->contents, $encoding));
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('url', $doc->url, $encoding));
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('date', $doc->date, $encoding));
     //additional data added on a per-module basis
     $this->addField(Zend_Search_Lucene_Field::Binary('data', serialize($data)));
     // adding a path allows the document to know where to find specific library calls
     // for checking access to a module or block content. The Lucene records should only
     // be responsible to bring back to that call sufficient and consistent information
     // in order to perform the check.
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('path', $path, $encoding));
     /*
     // adding a capability set required for viewing. -1 if no capability required.
     // the capability required for viewing is depending on the local situation
     // of the document. each module should provide this information when pushing
     // out search document structure. Although capability model should be kept flat
     // there is no exclusion some module or block developpers use logical combinations
     // of multiple capabilities in their code. This possibility should be left open here.
     $this->addField(Zend_Search_Lucene_Field::UnIndexed('capabilities', $caps));
     */
     /*
     // Additional key set allows a module to ask for extensible criteria based search
     // depending on the module internal needs.
     */
     if (!empty($additional_keyset)) {
         foreach ($additional_keyset as $keyname => $keyvalue) {
             $this->addField(Zend_Search_Lucene_Field::Keyword($keyname, $keyvalue, $encoding));
         }
     }
 }